TF-PROTACs Enable Targeted Degradation of Transcription Factors

Abstract

Transcription factors (TFs) represent a major class of therapeutic targets for the treatment of human diseases including cancer. Although the biological functions and even crystal structures of many TFs have been clearly elucidated, there is still no viable approach to target the majority of TFs, thus rendering them undruggable for decades. PROTACs (proteolysis targeting chimeras) emerge as a powerful class of therapeutic modalities, which rely on induced protein-protein interactions between the proteins of interest (POIs) and E3 ubiquitin ligases to aid the degradation of POIs by the ubiquitin-proteasome system (UPS). Here, we report the development of a platform termed TF-PROTAC, which links an DNA oligonucleotide to an E3 ligase ligand via a click reaction, to selectively degrade the TF of interest. The selectivity of these TF-PROTACs depends on the DNA oligonucleotides utilized that can be specific to the TFs of interest. We have developed two series of VHL-based TF-PROTACs, NF-κB-PROTAC (dNF-κB) and E2F-PROTAC (dE2F), which effectively degrade endogenous p65 and E2F1 proteins in cells, respectively, and subsequently display superior antiproliferative effects in cells. Collectively, our results suggest that TF-PROTACs provide a generalizable platform to achieve selective degradation of TFs and a universal strategy for targeting most "undruggable" TFs.

Figure 1.

A schematic diagram of the TF-PROTAC strategy. A BCN-modified VHL ligand (VHLL-X-BCN) is incorporated onto an azide-modified DNA oligomer (N3-ODN) via a copper-free strain-promoted azide–alkyne cycloaddition (SPAAC) reaction, forming a TF-PROTAC to recruit the VHL E3 ubiquitin ligase to ubiquitinate the transcription factor (TF) of interest, which is subsequently degradated by the 26S proteasome.

Figure 2.

In vitro click reaction that leads to the generation of TF-PROTACs to target NF-κB. (A) A schematic diagram for the NF-κB motif and N3-NF-κB-ODN. (B) N3-NF-κB was capable of binding with NF-κB subunit p65. (C) Incorporation of VHLL-BCN #1 onto the N3-NF-κB-ODN led to an increase of molecular weight of 664 Da, which can be clearly separated by 20% native PAGE. (D) The click efficiency between VHLL-BCN #1 and N3-NF-κB-ODN with different ratios. A mixture of VHLL-BCN #1 and N3-NF-κB-ODN (50 μM) was incubated in PBS at 37 °C for indicated time points, followed by separation via 20% native PAGE. (E) dNF-κB #1 competed with Biotin-NF-κB for binding to RelA/p65. (F) dNF-κB #1 induced the VHL-TF-PROTAC-p65 ternary complex formation, which was determined by a GST pulldown assay.

Figure 3.

dNF-κB promotes the targeted degradation of the NF-κB transcription factor in cells. (A) Design of a series of 18 BCN-modified VHL ligands with various linkers between BCN and the VHL ligand. (B) In vitro click of VHLL-X-BCN (#1 - #18, 150 μM) with N3-NF-κB-ODN (50 μM), which can be separated by native PAGE. (C) Western blots for p65 in HeLa cells after treated with 10 μg/mL of dNF-κB (#1 - #18) for 12 h. (D) dNF-κB (#4, #7, #15, #16, and #18) led to the ubiquitination of p65 in cells. HEK293T cells were transfected with Flag-p65 for 24 h and then transfected with 10 μg/mL of indicated dNF-κB for another 12 h, followed by Western blotting for Flag-p65 in Ni-NTA pulldown and whole cell lysis (WCL). (E) Comparison of proteomic changes after treatment with dNF-κB #16 or control in HeLa cells. Dotted lines indicate either 50% loss or 2-fold increase of the protein level (x axis) and p = 0.01 (y axis). (F) The proteasome inhibitor MG132 blocked the degradation of endogenous p65 by dNF-κB (#15 and #16) in HeLa cells. HeLa cells were transfected with 10 μg/mL of indicated dNF-κB and then treated with 10 μM MG132, followed by Western blotting for p65. (G) dNF-κB (#15 and #16, 10 μ/mL) repressed the proliferation of HeLa cells. (H—I) dNF-κB (#15 and #16, 10 μg/mL) reduced the tumorigenesis of HeLa cells in a colony formation assay. *: p < 0.05.

Figure 4.

Design and in vitro click of dE2F. (A) A schematic diagram for the E2F motif and the sequence of N3-E2F-ODN. (B) Annealing of double strain N3-E2F-ODN. (C) N3-E2F-ODN was capable of binding with E2F1 transcription factor. (D) In vitro click of VHLL-BCN (#1 - #18, 150 μM) with N3-E2F-ODN (50 μM), which can be separated by 20% native PAGE. (E) Western blotting for E2F1 in HeLa cells after being treated with 10 μg/mL of dE2F (#1 - #18) for 12 h.

Figure 5.

dE2F promotes degradation of the E2F1 transcription factor in HeLa cells. (A) dE2F #16 and #17 degraded E2F1 in HeLa cells in a concentration-dependent manner. HeLa cells were transfected with 10 or 25 μg/mL of dE2F #16 or #17 for 12 h, followed by Western blotting for E2F1. (B) dE2F led to ubiquitination of E2F1 in cells. HeLa cells were transfected with HA-E2F1 for 24 h, then treated with 25 μg/mL of dE2F #16 or #17 for 12 h, followed by Western blotting for HA-E2F1 in Ni-NTA pulldown and WCL. (C) The proteasome inhibitor MG132 blocked the degradation of endogenous E2F1 by dE2F #16 and #17 in cells. HeLa cells were transfected with 25 μg/mL indicated dE2F1 and then treated with 30 μM MG132, followed by Western blotting for E2F1. (D) dE2F1 (#16 and #17, 25 μg/mL) repressed the proliferation of HeLa cells. (E, F) dE2F (#16 and #17, 25 μg/mL) reduced the tumorigenesis of HeLa cells in a colony formation assay. **, ***: p < 0.01, p < 0.001.